In [ ]:
%matplotlib inline

DCGAN Tutorial

Author: Nathan Inkawhich <https://github.com/inkawhich>__

In [1]:
from __future__ import print_function
#%matplotlib inline
import argparse
import os
import random
import torch
import torch.nn as nn
import torch.nn.parallel
import torch.backends.cudnn as cudnn
import torch.optim as optim
import torch.utils.data
import torchvision.datasets as dset
import torchvision.transforms as transforms
import torchvision.utils as vutils
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.animation as animation
from IPython.display import HTML

# Set random seed for reproducibility
manualSeed = 999
#manualSeed = random.randint(1, 10000) # use if you want new results
print("Random Seed: ", manualSeed)
random.seed(manualSeed)
torch.manual_seed(manualSeed)
Random Seed:  999
Out[1]:
<torch._C.Generator at 0x7f807aeabb30>

Inputs

Let’s define some inputs for the run:

  • dataroot - the path to the root of the dataset folder. We will talk more about the dataset in the next section
  • workers - the number of worker threads for loading the data with the DataLoader
  • batch_size - the batch size used in training. The DCGAN paper uses a batch size of 128
  • image_size - the spatial size of the images used for training. This implementation defaults to 64x64. If another size is desired, the structures of D and G must be changed. See here <https://github.com/pytorch/examples/issues/70>__ for more details
  • nc - number of color channels in the input images. For color images this is 3
  • nz - length of latent vector
  • ngf - relates to the depth of feature maps carried through the generator
  • ndf - sets the depth of feature maps propagated through the discriminator
  • num_epochs - number of training epochs to run. Training for longer will probably lead to better results but will also take much longer
  • lr - learning rate for training. As described in the DCGAN paper, this number should be 0.0002
  • beta1 - beta1 hyperparameter for Adam optimizers. As described in paper, this number should be 0.5
  • ngpu - number of GPUs available. If this is 0, code will run in CPU mode. If this number is greater than 0 it will run on that number of GPUs
In [11]:
# Root directory for dataset
dataroot = "../data/raw/planctons_original"

# Number of workers for dataloader
workers = 10

# Batch size during training
batch_size = 128

# Spatial size of training images. All images will be resized to this
#   size using a transformer.
image_size = 64

# Number of channels in the training images. For color images this is 3
nc = 3

# Size of z latent vector (i.e. size of generator input)
nz = 100

# Size of feature maps in generator
ngf = 64

# Size of feature maps in discriminator
ndf = 64

# Number of training epochs
num_epochs = 20

# Learning rate for optimizers
lr = 0.0002

# Beta1 hyperparam for Adam optimizers
beta1 = 0.5

# Number of GPUs available. Use 0 for CPU mode.
ngpu = 1

Data

In [12]:
# We can use an image folder dataset the way we have it setup.
# Create the dataset
dataset = dset.ImageFolder(root=dataroot,
                           transform=transforms.Compose([
                               transforms.Resize(image_size),
                               transforms.CenterCrop(image_size),
                               transforms.ToTensor(),
                               transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5)),
                           ]))
# Create the dataloader
dataloader = torch.utils.data.DataLoader(dataset, batch_size=batch_size,
                                         shuffle=True, num_workers=workers)

# Decide which device we want to run on
device = torch.device("cuda:0" if (torch.cuda.is_available() and ngpu > 0) else "cpu")

# Plot some training images
real_batch = next(iter(dataloader))
plt.figure(figsize=(8,8))
plt.axis("off")
plt.title("Training Images")
plt.imshow(np.transpose(vutils.make_grid(real_batch[0].to(device)[:64], padding=2, normalize=True).cpu(),(1,2,0)))
Out[12]:
<matplotlib.image.AxesImage at 0x7f80791898e0>

Implementation

With our input parameters set and the dataset prepared, we can now get into the implementation. We will start with the weigth initialization strategy, then talk about the generator, discriminator, loss functions, and training loop in detail.

Weight Initialization ~~~~~

From the DCGAN paper, the authors specify that all model weights shall be randomly initialized from a Normal distribution with mean=0, stdev=0.02. The weights_init function takes an initialized model as input and reinitializes all convolutional, convolutional-transpose, and batch normalization layers to meet this criteria. This function is applied to the models immediately after initialization.

In [13]:
# custom weights initialization called on netG and netD
def weights_init(m):
    classname = m.__class__.__name__
    if classname.find('Conv') != -1:
        nn.init.normal_(m.weight.data, 0.0, 0.02)
    elif classname.find('BatchNorm') != -1:
        nn.init.normal_(m.weight.data, 1.0, 0.02)
        nn.init.constant_(m.bias.data, 0)
In [14]:
# Generator Code

class Generator(nn.Module):
    def __init__(self, ngpu):
        super(Generator, self).__init__()
        self.ngpu = ngpu
        self.main = nn.Sequential(
            # input is Z, going into a convolution
            nn.ConvTranspose2d( nz, ngf * 8, 4, 1, 0, bias=False),
            nn.BatchNorm2d(ngf * 8),
            nn.ReLU(True),
            # state size. (ngf*8) x 4 x 4
            nn.ConvTranspose2d(ngf * 8, ngf * 4, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ngf * 4),
            nn.ReLU(True),
            # state size. (ngf*4) x 8 x 8
            nn.ConvTranspose2d( ngf * 4, ngf * 2, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ngf * 2),
            nn.ReLU(True),
            # state size. (ngf*2) x 16 x 16
            nn.ConvTranspose2d( ngf * 2, ngf, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ngf),
            nn.ReLU(True),
            # state size. (ngf) x 32 x 32
            nn.ConvTranspose2d( ngf, nc, 4, 2, 1, bias=False),
            nn.Tanh()
            # state size. (nc) x 64 x 64
        )

    def forward(self, input):
        return self.main(input)

Now, we can instantiate the generator and apply the weights_init function. Check out the printed model to see how the generator object is structured.

In [15]:
# Create the generator
netG = Generator(ngpu).to(device)

# Handle multi-gpu if desired
if (device.type == 'cuda') and (ngpu > 1):
    netG = nn.DataParallel(netG, list(range(ngpu)))

# Apply the weights_init function to randomly initialize all weights
#  to mean=0, stdev=0.2.
netG.apply(weights_init)

# Print the model
print(netG)
Generator(
  (main): Sequential(
    (0): ConvTranspose2d(100, 512, kernel_size=(4, 4), stride=(1, 1), bias=False)
    (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (2): ReLU(inplace=True)
    (3): ConvTranspose2d(512, 256, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (4): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (5): ReLU(inplace=True)
    (6): ConvTranspose2d(256, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (7): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (8): ReLU(inplace=True)
    (9): ConvTranspose2d(128, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (10): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (11): ReLU(inplace=True)
    (12): ConvTranspose2d(64, 3, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (13): Tanh()
  )
)

Discriminator Code

In [16]:
class Discriminator(nn.Module):
    def __init__(self, ngpu):
        super(Discriminator, self).__init__()
        self.ngpu = ngpu
        self.main = nn.Sequential(
            # input is (nc) x 64 x 64
            nn.Conv2d(nc, ndf, 4, 2, 1, bias=False),
            nn.LeakyReLU(0.2, inplace=True),
            # state size. (ndf) x 32 x 32
            nn.Conv2d(ndf, ndf * 2, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ndf * 2),
            nn.LeakyReLU(0.2, inplace=True),
            # state size. (ndf*2) x 16 x 16
            nn.Conv2d(ndf * 2, ndf * 4, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ndf * 4),
            nn.LeakyReLU(0.2, inplace=True),
            # state size. (ndf*4) x 8 x 8
            nn.Conv2d(ndf * 4, ndf * 8, 4, 2, 1, bias=False),
            nn.BatchNorm2d(ndf * 8),
            nn.LeakyReLU(0.2, inplace=True),
            # state size. (ndf*8) x 4 x 4
            nn.Conv2d(ndf * 8, 1, 4, 1, 0, bias=False),
            nn.Sigmoid()
        )

    def forward(self, input):
        return self.main(input)

Now, as with the generator, we can create the discriminator, apply the weights_init function, and print the model’s structure.

In [17]:
# Create the Discriminator
netD = Discriminator(ngpu).to(device)

# Handle multi-gpu if desired
if (device.type == 'cuda') and (ngpu > 1):
    netD = nn.DataParallel(netD, list(range(ngpu)))
    
# Apply the weights_init function to randomly initialize all weights
#  to mean=0, stdev=0.2.
netD.apply(weights_init)

# Print the model
print(netD)
Discriminator(
  (main): Sequential(
    (0): Conv2d(3, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (1): LeakyReLU(negative_slope=0.2, inplace=True)
    (2): Conv2d(64, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (3): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (4): LeakyReLU(negative_slope=0.2, inplace=True)
    (5): Conv2d(128, 256, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (6): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (7): LeakyReLU(negative_slope=0.2, inplace=True)
    (8): Conv2d(256, 512, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)
    (9): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)
    (10): LeakyReLU(negative_slope=0.2, inplace=True)
    (11): Conv2d(512, 1, kernel_size=(4, 4), stride=(1, 1), bias=False)
    (12): Sigmoid()
  )
)
In [18]:
# Initialize BCELoss function
criterion = nn.BCELoss()

# Create batch of latent vectors that we will use to visualize
#  the progression of the generator
fixed_noise = torch.randn(64, nz, 1, 1, device=device)

# Establish convention for real and fake labels during training
real_label = 1
fake_label = 0

# Setup Adam optimizers for both G and D
optimizerD = optim.Adam(netD.parameters(), lr=lr, betas=(beta1, 0.999))
optimizerG = optim.Adam(netG.parameters(), lr=lr, betas=(beta1, 0.999))
In [19]:
# Training Loop

# Lists to keep track of progress
img_list = []
G_losses = []
D_losses = []
iters = 0

print("Starting Training Loop...")
# For each epoch
for epoch in range(num_epochs):
    # For each batch in the dataloader
    for i, data in enumerate(dataloader, 0):
        
        ############################
        # (1) Update D network: maximize log(D(x)) + log(1 - D(G(z)))
        ###########################
        ## Train with all-real batch
        netD.zero_grad()
        # Format batch
        real_cpu = data[0].to(device)
        b_size = real_cpu.size(0)
        label = torch.full((b_size,), real_label, device=device)
        # Forward pass real batch through D
        output = netD(real_cpu).view(-1)
        # Calculate loss on all-real batch
        errD_real = criterion(output, label)
        # Calculate gradients for D in backward pass
        errD_real.backward()
        D_x = output.mean().item()

        ## Train with all-fake batch
        # Generate batch of latent vectors
        noise = torch.randn(b_size, nz, 1, 1, device=device)
        # Generate fake image batch with G
        fake = netG(noise)
        label.fill_(fake_label)
        # Classify all fake batch with D
        output = netD(fake.detach()).view(-1)
        # Calculate D's loss on the all-fake batch
        errD_fake = criterion(output, label)
        # Calculate the gradients for this batch
        errD_fake.backward()
        D_G_z1 = output.mean().item()
        # Add the gradients from the all-real and all-fake batches
        errD = errD_real + errD_fake
        # Update D
        optimizerD.step()

        ############################
        # (2) Update G network: maximize log(D(G(z)))
        ###########################
        netG.zero_grad()
        label.fill_(real_label)  # fake labels are real for generator cost
        # Since we just updated D, perform another forward pass of all-fake batch through D
        output = netD(fake).view(-1)
        # Calculate G's loss based on this output
        errG = criterion(output, label)
        # Calculate gradients for G
        errG.backward()
        D_G_z2 = output.mean().item()
        # Update G
        optimizerG.step()
        
        # Output training stats
        if i % 50 == 0:
            print('[%d/%d][%d/%d]\tLoss_D: %.4f\tLoss_G: %.4f\tD(x): %.4f\tD(G(z)): %.4f / %.4f'
                  % (epoch, num_epochs, i, len(dataloader),
                     errD.item(), errG.item(), D_x, D_G_z1, D_G_z2))
        
        # Save Losses for plotting later
        G_losses.append(errG.item())
        D_losses.append(errD.item())
        
        # Check how the generator is doing by saving G's output on fixed_noise
        if (iters % 500 == 0) or ((epoch == num_epochs-1) and (i == len(dataloader)-1)):
            with torch.no_grad():
                fake = netG(fixed_noise).detach().cpu()
            img_list.append(vutils.make_grid(fake, padding=2, normalize=True))
            
        iters += 1
Starting Training Loop...
[0/20][0/29]	Loss_D: 2.2105	Loss_G: 3.6335	D(x): 0.3874	D(G(z)): 0.5888 / 0.0416
[1/20][0/29]	Loss_D: 2.4113	Loss_G: 19.1050	D(x): 0.2434	D(G(z)): 0.0000 / 0.0000
[2/20][0/29]	Loss_D: 0.7714	Loss_G: 17.4645	D(x): 0.9809	D(G(z)): 0.4036 / 0.0000
[3/20][0/29]	Loss_D: 0.3540	Loss_G: 3.8919	D(x): 0.7624	D(G(z)): 0.0366 / 0.0279
[4/20][0/29]	Loss_D: 0.7067	Loss_G: 3.0759	D(x): 0.5688	D(G(z)): 0.0194 / 0.0588
[5/20][0/29]	Loss_D: 0.3579	Loss_G: 2.8328	D(x): 0.7753	D(G(z)): 0.0822 / 0.0712
[6/20][0/29]	Loss_D: 0.1554	Loss_G: 5.4948	D(x): 0.8849	D(G(z)): 0.0201 / 0.0201
[7/20][0/29]	Loss_D: 0.3085	Loss_G: 3.7066	D(x): 0.7773	D(G(z)): 0.0298 / 0.0485
[8/20][0/29]	Loss_D: 0.5155	Loss_G: 6.4880	D(x): 0.9150	D(G(z)): 0.3292 / 0.0053
[9/20][0/29]	Loss_D: 0.1973	Loss_G: 4.2892	D(x): 0.9589	D(G(z)): 0.1363 / 0.0197
[10/20][0/29]	Loss_D: 0.3157	Loss_G: 2.3657	D(x): 0.9548	D(G(z)): 0.2255 / 0.1092
[11/20][0/29]	Loss_D: 0.6580	Loss_G: 2.9177	D(x): 0.8003	D(G(z)): 0.3307 / 0.0655
[12/20][0/29]	Loss_D: 0.5599	Loss_G: 2.3519	D(x): 0.7533	D(G(z)): 0.2209 / 0.1021
[13/20][0/29]	Loss_D: 1.5516	Loss_G: 1.3211	D(x): 0.2865	D(G(z)): 0.0163 / 0.3211
[14/20][0/29]	Loss_D: 0.4326	Loss_G: 4.0482	D(x): 0.7048	D(G(z)): 0.0343 / 0.0237
[15/20][0/29]	Loss_D: 0.5843	Loss_G: 2.9635	D(x): 0.8469	D(G(z)): 0.3066 / 0.0805
[16/20][0/29]	Loss_D: 0.9348	Loss_G: 2.0497	D(x): 0.7218	D(G(z)): 0.3800 / 0.1693
[17/20][0/29]	Loss_D: 0.5320	Loss_G: 2.2532	D(x): 0.8802	D(G(z)): 0.2947 / 0.1337
[18/20][0/29]	Loss_D: 0.3975	Loss_G: 2.9881	D(x): 0.8588	D(G(z)): 0.1841 / 0.0624
[19/20][0/29]	Loss_D: 0.5798	Loss_G: 2.4598	D(x): 0.6696	D(G(z)): 0.1229 / 0.1084

Results

Finally, lets check out how we did. Here, we will look at three different results. First, we will see how D and G’s losses changed during training. Second, we will visualize G’s output on the fixed_noise batch for every epoch. And third, we will look at a batch of real data next to a batch of fake data from G.

Loss versus training iteration

Below is a plot of D & G’s losses versus training iterations.

In [20]:
plt.figure(figsize=(10,5))
plt.title("Generator and Discriminator Loss During Training")
plt.plot(G_losses,label="G")
plt.plot(D_losses,label="D")
plt.xlabel("iterations")
plt.ylabel("Loss")
plt.legend()
plt.show()

Visualization of G’s progression

Remember how we saved the generator’s output on the fixed_noise batch after every epoch of training. Now, we can visualize the training progression of G with an animation. Press the play button to start the animation.

In [21]:
#%%capture
fig = plt.figure(figsize=(8,8))
plt.axis("off")
ims = [[plt.imshow(np.transpose(i,(1,2,0)), animated=True)] for i in img_list]
ani = animation.ArtistAnimation(fig, ims, interval=1000, repeat_delay=1000, blit=True)

HTML(ani.to_jshtml())
Out[21]:

Real Images vs. Fake Images

Finally, lets take a look at some real images and fake images side by side.

In [22]:
# Grab a batch of real images from the dataloader
real_batch = next(iter(dataloader))

# Plot the real images
plt.figure(figsize=(15,15))
plt.subplot(1,2,1)
plt.axis("off")
plt.title("Real Images")
plt.imshow(np.transpose(vutils.make_grid(real_batch[0].to(device)[:64], padding=5, normalize=True).cpu(),(1,2,0)))

# Plot the fake images from the last epoch
plt.subplot(1,2,2)
plt.axis("off")
plt.title("Fake Images")
plt.imshow(np.transpose(img_list[-1],(1,2,0)))
plt.show()

Where to Go Next

We have reached the end of our journey, but there are several places you could go from here. You could:

  • Train for longer to see how good the results get
  • Modify this model to take a different dataset and possibly change the size of the images and the model architecture
  • Check out some other cool GAN projects here <https://github.com/nashory/gans-awesome-applications>__
  • Create GANs that generate music <https://deepmind.com/blog/wavenet-generative-model-raw-audio/>__